Research map / Architectures and efficiency / Attention and transformers
Improving Context-Aware Neural Machine Translation Using Self-Attentive Sentence Embedding
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| hierarchical self-attentive context encoder (HCE) instead of concatenating context (TwC) | Transformer · Web-crawled En-Ko | BLEU: 9.46 -> 11.30 | Helped |
| HCE vs Transformer without contexts | Transformer · IWSLT'17 En-De | BLEU: 28.25 -> 28.89 | Helped |
| HCE vs HAN | context-aware Transformer · OpenSubtitles En-De | BLEU: 28.00 -> 28.40 | Helped |
| HCE vs Transformer without contexts on OpenSubtitles En-Tr | Transformer · OpenSubtitles En-Tr | BLEU: 24.89 -> 25.11 | Helped |
| concatenating context sentences (TwC) vs no context | Transformer · OpenSubtitles En-Tr | BLEU: 24.89 -> 23.96 | Hurt |
| add DCT context encoder to Transformer | Transformer · IWSLT'17 En-De | BLEU: 28.25 -> 26.76 | Hurt |
| HCE vs baselines on helpful vs unhelpful BLEU gap | En-Ko translation · En-Ko helpful vs unhelpful gap | BLEU gap: +1.57 (HAN) -> +0.32 | Helped |
| hierarchical context encoding (HCE) vs concatenation (TwC) | Transformer · En-Ko pronoun resolution suite | accuracy: 0.25 -> 0.48 | Helped |
| HCE vs HAN on pronoun resolution | context-aware Transformer · En-Ko pronoun resolution suite | accuracy: 0.47 -> 0.48 | Helped |
| HCE vs TwC training speed | Transformer · Training speed | steps/sec: 4.07 -> 4.67 | Helped |
Closest papers by meaning
- Context-Aware Self-Attention Networks (AAAI 2019 · similarity 0.87)
- Syntax-Directed Attention for Neural Machine Translation (AAAI 2018 · similarity 0.87)
- Core Context Aware Transformers for Long Context Language Modeling (ICML 2025 · similarity 0.85)
- A Paradigm Shift in Machine Translation: Boosting Translation Performance of Large Language Models (ICLR 2024 · similarity 0.85)
- Hierarchical Context Merging: Better Long Context Understanding for Pre-trained LLMs (ICLR 2024 · similarity 0.85)
- Quantifying the Plausibility of Context Reliance in Neural Machine Translation (ICLR 2024 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.